Salam International - Essay Example

Salam International December 27, Salam International Determining the underlying relationship in any model is important asit determines the suitability of any given model to not only adequately representing the data, but also provision of accurate estimates of the parameter estimates. Correlation and regression estimates have proved useful tools in data analysis for describing the relationship between the dependent and independent variables. While the correlation analysis seeks to measure the strength of relationship between the dependent and independent variables in the model, regression analysis on the other hand seeks to determine how changes in independent variables affect the dependent variable (Archdeacon, 1994).

From the excel output, the correlation matrix provides that while the correlation coefficient (R2) is 0.109, the adjusted correlation coefficient is -0.010 and that of the independent P/E and P/B ratios is -0.0102. The result of the correlation coefficient implies that 10.9% of the total variation is explained by the independent variables P/E and P/B while the remaining 89.1% represents the unexplained variation by the regression model. Consequently, the result of the adjusted R2 implies that the model implies that the model presents the best estimate of the degree of the relationship for our variables in the population under study.

Consequently, the output presents the standard error as 0.0255 which implies that approximately 95% of the observations should fall within plus/minus 5% of the of the line fitted by the model which presents a close match for the prediction interval. Lastly, the intercept of 0.067 implies that given a P/B ratio of 0, then we can predict ROE as 0.067.Therefore, we can present the regression model as:Y = 0.067 – 0.039XWhere Y – The dependent variable (ROE) X – Independent variable (P/B)The slope of the regression curve is 0.

067 which implies that an increase in P/B ratio by 1 results to a decrease in ROE by approximately 0.039.The correlation matrix from the correlation analysis postulate that the correlation of P/B ratio and ROE is -0.324 which indicates that the correlation is weak and negative hence a change in one variable results to a change in the opposite direction of the other variables. Consequently, the correlation between P/E ratio and ROE is -0.226 which implies a weak negative correlation which indicates that a change in one variable to a large extent does not determine the direction of change of the other variable.

The correlation of the P/E and P/B ratios is 0.528 which implies a strong positive correlation the change in one variable, either P/E or P/B ratio results to an equal change of the other variable to the same direction hence an increase in P/E results to an increase in P/B and a decrease in P/E results to a decrease in P/B and the inverse is also true. T-statistics seeks to measure the likelihood of the actual value of the parameter not being zero (Archdeacon, 1994). From the excel output, the t-statistics values for the intercept is 2.

264 while that of the P/B.V Ratio and P/E ratio is -0.987 and -0.266 respectively. The P value on the other hand presents the probability of getting the t-statistics value from the data. The P-values from the excel output indicate values as 0.039, 0.339 and 0.794 for Intercept, P/B.V Ratio and P/E Ratio respectively. From the values of P and t-stat, we can conclude that we are 3.9%, 33.9% and 79.4% that the t-static values for the population are as indicated.ReferencesArchdeacon, T. J. (1994).

 Correlation and regression analysis: A historians guide.Madison: University of Wisconsin Press.